A 9:16 File Is Not a Placement Strategy: Buy Vertical Video by Viewing Environment

The procurement answer is simple: do not commission near-identical “TikTok,” “Reels,” “Shorts,” and “programmatic vertical” edits by default. But do not push one universal 9:16 export into every available placement either.

A better unit of purchase is one fact-stable vertical master plus a small number of viewing-environment branches.

That distinction matters more after Google’s September 28, 2026 launch of Vertical Video Unification in Display & Video 360. Google says DV360 can now centralize the programmatic buying and measurement of vertical video across channels. In the initial Hellmann’s and WPP Media test disclosed by Google, unique reach increased by 24% while cost per unique user fell by 25%.

Those numbers describe one early test reported by the platform. They are not a universal benchmark. The structural signal is the workflow change: media teams can increasingly manage vertical inventory through one buying and measurement layer.

Creative teams should not confuse that unification with sameness.

The buying interface is converging; the viewing experience is not

On the same date, IAB published A New Standard for Classifying Video Advertising. Its most useful line for a production brief is that a platform is not an attribute. One platform can distribute inventory across multiple media types, depending on the surface, device and delivery context.

The IAB framework separates three broad viewing environments—large-screen communal viewing, personal screens and public ambient displays—then adds operational signals such as sound, skippability, screen presentation, identity, content context, device, ad format, screen size and supply type.

That is a better creative model than a list of platform logos. A vertical ad may appear:

• in a personal feed where the viewer can scroll immediately;
• full-screen inside a streaming or publisher app;
• beside editorial copy, a product card or another interface;
• with sound on by default, sound off by default or no dependable audio state;
• with a click-through, a product action or no immediate conversion path.

The file may still be 9:16 in every case. The viewer’s relationship with it is not.

Google’s current Display & Video 360 creative guidance accepts portrait video in 9:16 or 3:4. Its video creative documentation also explains that transcodes preserve the source orientation. That solves technical delivery. It does not decide whether the opening survives a fast scroll, whether product proof remains visible under interface elements, or whether a call to action matches the destination.

The ONCE Vertical Environment Matrix

Before commissioning “six short videos,” describe each intended environment in six fields.

1. Viewing mode — personal screen, communal screen or public display; continuous viewing or rapid feed exposure.
2. Sound state — on by default, off by default or uncertain; whether the essential claim works through picture and captions alone.
3. Exit mechanic — skip button, scroll, close control or completion-required delivery; what the first two seconds must accomplish.
4. Screen presentation — full-screen, in-feed, picture-in-picture or paired with product UI; where faces, products and captions can safely sit.
5. Action path — continue watching, visit a page, open a product, request information or add to cart; whether the ending accurately hands off to that action.
6. Rights and measurement — music, talent, creator and territory permissions; asset ID, branch ID and activation record.

If only the platform name changes while these six conditions remain substantially the same, one branch may be enough. If two or more of sound, exit mechanic, screen presentation and action path change materially, create a different environment branch.

If the product proof itself cannot survive the safe composition—for example a small connector, interface state or hand movement must sit at the edge of the frame—the answer is not another caption pass. Capture an alternative angle during the shoot or schedule a pickup.

This is the same operational lesson explored in our analysis of Amazon’s full-funnel automation: the more a media system can assemble and allocate assets, the more precisely the production package must state what each unit can do.

What belongs in the master, and what should branch

The vertical master is not the final ad. It is the controlled production base from which ads can be derived.

Keep these elements stable in the master package:

• approved product facts and proof shots;
• core brand identifiers and visual behavior;
• clean dialogue, music and ambience stems;
• caption text and textless plates;
• project files and source footage;
• talent, music, creator, market and term rights;
• a stable master asset ID.

Change only what the viewing environment requires:

• Opening order: lead with the product action or problem in a scrollable feed; allow more context when viewing is deliberate.
• Information redundancy: make picture and captions independently useful in uncertain-audio environments; avoid making narration and large text compete when sound is dependable.
• Shot duration: hold small-screen product proof closer and longer; preserve fuller actions in higher-attention, full-screen contexts.
• End action: send commerce inventory to a product action, brand inventory to deeper understanding, and B2B inventory to a relevant information or inquiry step.
• Asset record: attach the master ID, branch ID, intended environment, rights window and edit history to each activated file.

For a short-lived campaign in one market, one personal-screen feed and one destination, a master and a single branch may be sufficient. If the plan spans social feeds, publisher vertical inventory, streaming apps and commerce surfaces, build the matrix before shooting. Discovering those differences in post often means recutting dialogue, rebuilding captions, searching for missing proof or scheduling a pickup.

Stop pricing by platform version count

The number of platforms is a poor cost denominator. It can count the same edit several times while hiding the work that actually makes an asset deployable.

Use a transparent project equation instead:

Total vertical-video investment = master capture and proof assets + environment branch edits + technical and visual QA + rights extensions + asset labeling and measurement preparation

Then compare suppliers or scopes with two ratios:

• Cost per qualified viewing environment = total investment ÷ branches that pass review and can actually be activated.
• Cost per reusable proof unit = master and proof-production investment ÷ approved fact units that can safely serve more than one branch.

No invented industry average is required. Compare two real scopes: a platform-by-platform package and a master-plus-environment package. The second is more scalable only if it reduces repeated capture while preserving the creative conditions, rights and handoff required by each environment.

Copying one export into more inventory increases distribution. It does not automatically increase creative coverage.

Four acceptance questions before activation

For every branch, ask:

1. Without sound, can a viewer understand the product and the problem in the first two seconds?
2. Under interface overlays or partial-screen presentation, are the critical action, caption and brand still visible?
3. Does the next step after the video match the promise made inside it?
4. Do the music, talent, creator and product permissions cover this market, term and media environment?

Passing resolution and duration checks makes a file uploadable. Passing these four questions makes it closer to deployable.

Vertical media buying is becoming easier to unify. That is precisely why production briefs must become more explicit about context. For brands connected to Chinese supply chains and selling across markets, the useful next step is to build a reusable master through ONCE’s cross-border content workflow and branch it by real viewing conditions—not by the name of the app printed on the media plan.